the Relationship Between Activation Outliers and Feature Death
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
On the Relationship Between Activation Outliers and Feature Death in Sparse Autoencoders
arXiv:2605.31518v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) decompose neural network activations into interpretable features, but many learned features never activate, a problem called feature death that wastes dictionary capacity and can reintroduce superposition. Death rates vary dramatically between models: near-zero on GPT-2, over 70% on AlphaFold3 with identical configurations. We find that dimension-level activation outliers (dimensions whose mean magnitude is large...
AI Job Grief: The Unnamed Psychological Crisis Hitting Tech Workers
AI Job Grief: The Unnamed Psychological Crisis Hitting Tech Workers In the summer of 2025, an Epic Games layoff cut a worker who was a terminally ill father. According to the most-discussed account of the episode, his family lost his life insurance along with the job.